The HyperProbe project proudly announces its collaboration with NVIDIA, leveraging their cutting-edge Clara AI platform. This collaboration marks a significant step forward in medical imaging research and development, enhancing our capabilities and accelerating progress towards innovative solutions in the HypeRrobe project.
NVIDIA’s Clara platform offers a versatile toolkit tailored for medical imaging initiatives, bridging the gap from concept to production efficiently. For more information on the benefits of using the Clara platform, visit NVIDIA’s website here.
Within the HyperProbe project, Clara’s capabilities will be harnessed across multiple work packages, notably Work Packages 3, 4, and 5.
Work Package 3: Optical and Metrological Characterization
Led by UCL, this Work Package focuses on optical and metrological characterization. This involves rigorous measurement and analysis to understand the performance and capabilities of the HyperProbe device.
Utilizing Clara workstations, researchers at UCL will design and refine phantoms—an essential component in medical imaging—to evaluate and tune device performance.
Advanced Monte Carlo simulations will be conducted using tools like Monte Carlo eXtreme (MCX) and Mesh Monte Carlo (MMC), providing invaluable insights into light propagation and interaction within the phantom.
Angelos Artemiou, PhD student at UCL with the NVIDIA Clara device.
Work Package 4: End-to-End Trainable Pipeline for Image Reconstruction and Analysis
NVIDIA Clara arrived to the Technical University of Munich.
At the Technical University of Munich, the emphasis is on machine learning development to establish pixel-wise mappings between hyperspectral imaging data and molecular tissue profiles.
This approach, distinct from traditional supervised learning methods, leverages physical models to infer molecular composition from light spectra reflected off tissue surfaces.
NVIDIA’s parallel processing capabilities enable real-time optimization, essential for clinical applications requiring instantaneous feedback during surgical procedures.
Work Package 5: Validation Against Gold Standards
In this phase, the HyperProbe project validates its findings against established gold standards, particularly in intraoperative optical imaging for neurosurgical procedures.
Previous studies have demonstrated the feasibility of real-time identification of cortical activation areas using RGB imaging combined with quantitative modeling of brain hemodynamics.
Integration of Clara’s developer kit enhances the project’s ability to develop and validate real-time software for identifying functional brain areas, with potential applications in surgical settings.
The CREATIS team Université Claude Bernard Lyon 1 also received NVIDIA Clara.
The integration of NVIDIA’s Clara platform into the HyperProbe project represents a significant milestone in the project. By harnessing the power of AI and advanced computing technologies, the project aims to revolutionize neurosurgical procedures, assisting surgeons in accurately localizing cancerous and healthy cells during operations. We are grateful for NVIDIA’s collaboration and look forward to the impactful outcomes this partnership will bring to the field of medical imaging.
